In this case, the RDSH server performance degrades over time as more users log in. The following graph shows readings for users working remotely in Microsoft Word. Now let's look at what you'll see in the report if performance for an app is degraded. (You can get the same information in Perfmon by switching to the Report graph type.) Type of counter This table shows a visual example of these instances. In addition, there are two instances called "Max" (the maximum user input delay across all sessions) and "Average" (the average acorss all sessions). There are instances for each session ID, and their counters show the user input delay of any process within the specified session. Next, let's look at the User Input Delay per Session. Note that the maximum scale is set to 100 (ms) by default.
The counter starts reporting user input delay as soon as you add it. You won't see any processes that are running as SYSTEM. To use these new performance counters, you must first enable a registry key by running this command: reg add "HKLM\System\CurrentControlSet\Control\Terminal Server" /v "EnableLagCounter" /t REG_DWORD /d 0x1 /f NumberĮnable and use the new performance counters The counter reports the slowest user input delay in the interval because the user's perception of "slow" is determined by the slowest input time (the maximum) they experience, not the average speed of all total inputs. This is the longest time it takes for an input to reach the application, which can impact the speed of important and visible actions like typing.įor example, in the following table, the user input delay would be reported as 1,000 ms within this interval. One important detail of this counter is that it reports the maximum user input delay within a configurable interval. The User Input Delay counter measures the max delta (within an interval of time) between the input being queued and when it's picked up by the app in a traditional message loop, as shown in the following flow chart: The following image shows a rough representation of user input flow from client to application. This counter measures how long any user input (such as mouse or keyboard usage) stays in the queue before it is picked up by a process, and the counter works in both local and remote sessions. The User Input Delay counter can help you quickly identify the root cause for bad end user RDP experiences. The User Input Delay counter is only compatible with: